Click the exhibit button below. Given the scheduler-policy (below), how much bandwidth can the best-effort traffic receive (PIR and CIR) when the ingress rate of each queue is 10Mbps?

Question

Exhibit

Options

• APIR = 4Mbps. CIR = 1Mbps
• BPIR = 1Mbps, CIR = 1 Mbps
• CPIR = 10Mbps, CIR = 1Mbps
• DPIR = 5Mbps, CIR = 1.25Mbps
• EPIR = 1.25Mbps, CIR = 1.25Mbps

• A
  2% (1)
• B
  79% (37)
• C
  4% (2)
• D
  13% (6)
• E
  2% (1)

Explanation

Option B is correct because the scheduler policy sets a hard PIR cap on best-effort traffic equal to its CIR - both at 1Mbps - meaning the queue cannot burst beyond its committed rate regardless of available bandwidth or ingress rate. Even though each queue ingresses at 10Mbps, the policy enforces a strict ceiling: best-effort gets exactly 1Mbps and no more.

Why the distractors fail:

• A (PIR=4Mbps): Implies best-effort can burst above its CIR to 4Mbps, but the policy's PIR cap prevents any excess borrowing.
• C (PIR=10Mbps): Suggests best-effort could absorb the full ingress rate, which would only be true if no PIR cap were configured.
• D (PIR=5Mbps, CIR=1.25Mbps): These values are arithmetically tempting (e.g., dividing total bandwidth by queue count) but don't reflect what the policy actually configures.
• E (PIR=1.25Mbps): Same miscalculation error as D - deriving rates by simple division rather than reading the explicit policy values.

Memory tip: When PIR = CIR for a queue, think "no burst allowed" - the scheduler has locked the door at the committed rate. If PIR > CIR, the queue can borrow; if PIR = CIR, it's a hard ceiling with zero headroom.

No community discussion yet for this question.